Loop Engineering: Design Systems That Prompt Your AI Agents

Loop Engineering is an open source library of patterns, templates, and CLI tools for creating systems that orchestrate AI code agents like Claude Code, Codex, Cursor, etc. It’s based on the concept of “building a loop” instead of manually prompting the agent each time, your system periodically prompts the agent using persistent state, budgets, and skill files. Using the loop init CLI, you can set up a new loop within 5 minutes by generating STATE.md, LOOP.md files, budget and constraint files and get a Loop Ready Score. Use the loop doctor CLI to see the top 3 things you can do next. Other CLI tools include loop-audit (to get your Loop Ready Score), loop-cost (to estimate your token spend), loop-sync (to check for drift), loop-context (to manage memory), loop-worktree (to create isolated git worktrees), loop-gate (to enforce path denylists), and an MCP server (to lookup patterns). As Boris Cherny (Head of Claude Code at Anthropic) puts it: “I no longer prompt Claude. I have loops that prompt Claude and figure out what to do.” Introduced just 2 months ago, it has more than 10,500+ stars.

Set up loop-engineering for this project: https://github.com/cobusgreyling/loop-engineering — run npx @cobusgreyling/loop init . --pattern daily-triage --tool claude, then npx @cobusgreyling/loop doctor . and show me the Loop Ready Score and top 3 actions to improve it.

💻 Runs on a regular laptop — only Node.js needed.